In Phase 1, the enzymatic reduction step for stereoselective synthesis of chiral vicinal aminoalcohols was established. Eight different ketoreductases were evaluated, including 5 ketoreductases that were cloned and expressed during this project, and the enzymes were characterized for substrate and stereochemical preferences. Reduction of 2-substituted-beta-ketoesters, the key precursors for chiral aminoalcohols, gave products of high enantioselectivity and diastereoselectivity. Recycle numbers for nicotinamide cofactors of more than 5000 were achieved, leading to favorable economic projections. In Phase 2 we will create a library of stereoselective ketoreductase enzymes by mutagenizing cloned genes, enabling the stereoselective reduction of beta-ketoesters spanning a broad structural range. A cloning host will be developed that eliminates the need for antibiotic resistance for plasmid maintenance, and production of ketoreductase enzymes will be demonstrated in laboratory-scale fermentors. Reaction conditions will be optimized using two-phase aqueous-organic systems to maximize volumetric productivity for ketones that have limited solubility in water. The chemical rearrangement to produce the aminoalcohol products will be demonstrated, and a procedure based on the use of metal-affinity resins will be developed for recovering the aminoalcohol products from the reaction milieu. Three commercially important chiral vicinal aminoalcohols will be produced in multi-gram amounts to demonstrate the overall technology. PROPOSED COMMERCIAL APPLICATION: Production of key intermeidates for the production of existing and new pharmaceutical products.